Sm-Nd isotope and trace element study of Late Proterozoic metabasalts (“spilites”) from the Central Barrandian domain (Bohemian Massif, Czech Republic)
Published:January 01, 2007
- PDF LinkChapter PDF
Christian Pin, Jarmila Waldhausrová, 2007. "Sm-Nd isotope and trace element study of Late Proterozoic metabasalts (“spilites”) from the Central Barrandian domain (Bohemian Massif, Czech Republic)", The Evolution of the Rheic Ocean: From Avalonian-Cadomian Active Margin to Alleghenian-Variscan Collision, Ulf Linnemann, R. Damian Nance, Petr Kraft, Gernold Zulauf
Download citation file:
On the basis of immobile trace elements and Nd isotope signatures, the Barrandian meta-basalts may be ascribed to two major groups, extracted from contrasting mantle sources:
A depleted group, with strong light rare earth element depletion, elevated Zr/Nb ratios (>30), and highly radiogenic Nd isotopes (ϵNd600 from +7.8 to + 9.3). Multi-element patterns normalized to normal mid-ocean ridge basalt all show negative anomalies of Nb, and to a lesser degree, Zr and Ti. Eight samples may define a 605 ± 39-Ma whole-rock isochron with ϵNdi of +8.8 ± 0.2.
An enriched group, comprising both mildly enriched (Zr/Nb 12–18) and strongly enriched (Zr/Nb 4–7) samples, with ϵNd600 ranging from +8.2 to +3.8.
The depleted group is interpreted to reflect generation from depleted mantle sources fluxed by subduction-related components, probably in an intraoceanic back-arc basin. In contrast, the younger enriched group is typical of the within-plate style of mantle enrichment and documents the extinction of the subduction-related component. The switch from suprasubduction zone to within-plate magmatism suggests that new mantle material flowed into the former arc and back-arc system sources. This flow might have occurred simply as a result of ocean-ward migration of the subduction zone. Alternatively, the subduction fluxing might have stopped as a result of impingement of a spreading ridge with the intraoceanic trench, leading to mutual annihilation, a switch to a transform plate boundary, and opening of a slab window that allowed the inflow of new mantle and the generation of late-stage, within-plate enriched basalts. In terms of modern analogues, the Neoproterozoic of the Barrandian and other Cadomian regions of western Europe resemble arc and back-arc systems from the western Pacific region, where large intraoceanic subduction systems fringe major continental masses with a complex mosaic of microplates and magmatic arcs, including intervening basins floored either by oceanic crust or attenuated continental crust.
The geological record of early stages of accretion and evolution of continental lithosphere in western and central Europe was largely obscured, or even erased, by extremely strong tectonometamorphic overprinting, which occurred during the Variscan (Hercynian) cycle, from Cambro-Ordovician continental breakup to Carboniferous continent-continent collisional orogeny. For this reason, the few large, coherent domains that escaped pervasive reworking during Paleozoic times deserve special interest. Besides the Ossa-Morena zone of southwest Iberia and the Cadomian block of north Brittany and Normandy (France) in western Europe, the Teplá-Barrandian unit and the Bruno-Vistulicum block in central Europe provide the best examples to investigate well-preserved Late Proterozoic igneous and sedimentary formations and to gain insight into how and when the pre-Variscan lithosphere was built.
In the central part of the Bohemian Massif, the Teplá-Barrandian unit (Fig. 1) is composed of a low- to very-low-grade unit separated by major Variscan shear zones from the highly metamorphosed rocks of the surrounding Saxo-Thuringian segment of northwest vergence in the north, and Moldanubian segment of southeast vergence in the south (e.g., Glasmacher et al., 2002, and references therein). The Teplá-Barrandian block, and particularly its southeastern part (Barrandian area) displays a Precambrian basement, deformed and weakly metamorphosed during the Cadomian orogeny, and unconformably overlain by unmetamorphosed siliciclastic and carbonate sedimentary rocks of Cambrian to Middle Devonian age. This Paleozoic cover was only mildly deformed as a broad synclinorium prior to the deposition of Upper Carboniferous and Lower Permian continental sequences. Owing to these favorable circumstances, the Barrandian formations offer an excellent opportunity to study a relatively undisturbed segment of Late Precambrian crust, which might be representative of the much broader domain involved in, and strongly reworked by, the Cadomian and Variscan tectonometamorphic events. In particular, the nature of the materials forming the Late Precambrian crust (i.e., ancient crystalline continental basement versus juvenile crust and/or sediments) is pivotal to any model aiming to interpret the growth and evolution of European lithosphere.
The Neoproterozoic sedimentary pile of the Barrandian unit contains relatively abundant mafic metavolcanic rocks, often referred to as “spilites” (e.g., Fiala, 1977). Insofar as their original features and mantle source can be deciphered through low-grade alteration processes, these rocks may provide interesting clues to the geotectonic setting that prevailed during the formation of the Barrandian basin prior to the Cadomian orogeny and complement insights gained from the study of sedimentary country-rocks (e.g., Jakeš et al., 1979; Drost et al., 2004). In this work, the Sm-Nd isotope system, together with a set of trace elements selected on the basis of their relatively immobile behavior during postmagmatic processes, were used to reassess the geochemical characteristics of these igneous rocks and place constraints on their inferred mantle sources. The possible geodynamic implications of these results are explored in the broader scope of the European Cadomian orogenic domain.
Detailed information and interpretations on the evolution of the Teplá-Barrandian unit, including correlations with other Cadomian segments, can be found in a recent synthesis offered by Kříbek et al. (2000). Briefly, the Precambrian of the Central Barrandian domain (e.g., Chaloupsky et al., 1995) consists mostly of a very thick (probably >10 km), monotonous sequence of siliciclastic hemipelagites and turbidites. The basement is unknown, but seismic reflection data (9HR profile; Tomek and Dvořáková, 1994; Tomek et al., 1997) revealed different structures for the upper and lower parts of the Barrandian crust. The upper part is characterized by reflectors with variable positions and dips, in line with the mild Cadomian folding described by Holubec (1995a,b). In contrast, below ∼10 km, the 9HR profile imaged SSE-dipping multiple reflectors imbricated down to a depth of 20–25 km. This imbricate structure might be interpreted as an early Cadomian complex modified to some extent during the Variscan orogeny (Tomek et al., 1997), the reflectors possibly corresponding to individual thrust planes within a subduction-accretion complex similar to those found in modern arcs.
Following Kettner's pioneering work, most lithostratigraphic schemes proposed for this domain are based on the presence or absence of metabasalts, variably transformed into so-called “spilites.” According to recent proposals (e.g., Mašek, 2000), three major units can be distinguished, from bottom to top: (1) the Blovice Formation (∼7000 m thick?), containing frequent occurrences of mafic effusive igneous rocks, interbedded with shales, siltstones, graywackes, and minor rock-types, including pyroclastics, black shales, black cherts, and rare carbonates (diamictites are also reported to occur); (2) the Davle Formation (∼2000 m), composed of graywackes and shales associated with intermediate to felsic volcanics and volcaniclastics; and (3) the Štěchovice Group, composed almost exclusively of a thick pile (up to ∼5000 m) of flyschlike clastic rocks, including conglomerates, devoid of siliceous rocks, but containing alkaline volcanics. The Blovice and Davle formations are gathered into the Kralupy-Zbraslav Group, broadly corresponding to the “Pre-spilitic” and “Spilitic” of former schemes, whereas the Štěchovice Group corresponds to the “Post-spilitic series.” It should be stressed, however, that the stratigraphic contact between the Davle and Blovice formations was never observed, and that it would be preferable to establish different lithostratigraphic schemes in particular subdomains of the Barrandian Neoproterozoic (Röhlich, 2000). According to new mapping (Holubec 1995b), the Barrandian Neoproterozoic can be divided into three lithostratigraphic groups (Fig. 1): the Lower Group consists of the Rabštejn and úslava groups; the Middle, Zvíkovec Group; and the Upper, Rakovník Group.
Although the structure of the deep part of the Barrandian Neoproterozoic can only be inferred from seismic reflection profile 9HR, the near-surface geology displays relatively mild regional deformation, with three generations of folds and thrusts (Holubec, 1995a). An unconformity has been proposed but not fully demonstrated between the stratigraphically higher Zvíkovec and Rakovník groups (Holubec, 1995a). The corresponding tectonic phase would have caused shortening of the lower part of the Barrandian package.
Our study is focused on mafic volcanics of the Blovice Formation (sensu Mašek, 2000) in the central and northwestern part of the Barrandian region and does not deal with the Davle Formation, exposed only in the southeastern part of the Barrandian domain. The alkaline volcanics of the Rakovník Group (Holubec, 1995b), a lithostratigraphic equivalent of the Štěchovice Group occurring in the northwest flank of the Barrandian, were also studied.
Although a deep marine environment is inferred for the deposition of siliciclastic sediments of the Blovice Formation, very shallow conditions (lagoons, evaporitic flats) were locally achieved near the edges of volcanic islands, as shown by oolithic and stromatolitic carbonates (sometimes silicified) and pseudo-morphs after gypsum (e.g., Skoček and Pouba, 2000; Vavrdová, 2000) closely associated with mafic igneous rocks. Based on microfossils of cyanobacterial and algal origin, along with acritarchs, a Late Neoproterozoic (Vendian) age is ascribed to these sediments (e.g., Konzalová, 1980, 2000; Pacltová, 2000; Vavrdová, 2000). In the framework of recent geological time scale, these formations might therefore be considered as belonging to the Ediacaran system, spanning the ca. 630–542-Ma period (Gradstein et al., 2004). The age of deposition of the Štěchovice Group is constrained to be younger than ca. 570 Ma by isotopic dilution thermal ionization mass spectrometry (ID-TIMS) U-Pb ages measured for two rhyolitic boulders from conglomerates (585 ± 7 and 568 ± 3 Ma; Dörr et al., 2002) and for the youngest detrital zircons from a graywacke (564 ± 16 Ma by sensitive high-resolution ion microprobe [SHRIMP] U-Pb dating; Drost et al., 2004).
Important intercalations of predominantly mafic volcanic rocks, including frequent pillow lavas (Fiala, 1977) and often associated with metal-rich black shales (Pasava, 2000), occur within the Neoproterozoic sedimentary pile. These volcanics are exposed along several major belts, broadly parallel to the northeast–southwest-trending Variscan structures: the Svojšín belt and the Stříbro-Plasy volcanic belt in the northwest, the main central volcanic belt in the central Barrandian, and the southern volcanic zone (Fiala, 1977) in the southeast (Fig. 1). The samples analyzed in this study were collected from the central and northwestern volcanic belts and from the alkaline volcanics occurring in the flyschlike sediments of the Rakovník Group.
Based on previous studies (Fiala, 1977, 1978; Pelc and Waldhausrová, 1994; Waldhausrová, 1997a), three magmatic suites have been defined among the Neoproterozoic volcanics from the central and western Barrandian (Blovice Formation): (1) a primitive, tholeiitic suite, occurring in the lower part of the stratigraphic pile; (2) an alkaline suite, found in the upper part of the pile and representing the youngest Neoproterozoic volcanics; and (3) a chemically transitional suite, stratigraphically lying between the tholeiitic and alkaline suites. The major part of the volcanics were affected by prehnite-pumpellyite metamorphic-facies overprinting prior to the deposition of Cambrian cover sediments (Bernardová and Chab, 1974), but metamorphic grade increases to greenschist- to amphibolite-facies conditions toward the southwest, northwest, and north. A petrographic description, including electron microprobe analyses of the main relictual igneous phases and metamorphic minerals of the Barrandian metavolcanics, is given by Waldhausrová (1997a).
Twenty-two samples of meta-igneous rocks from the northwest, central, and southern volcanic zones have been selected for major and trace element chemistry and Sm-Nd isotopes (see Appendix for sample locations). All but one sample (Si-2, corresponding to a trachytic lava) are of broadly basaltic composition, reflecting the overwhelming proportion of mafic rocks in this part of the Barrandian area. In addition two samples of metagray-wackes (Viš-1 and Nml-1) and one of black shale (Kruš-2) have also been analyzed to get a crude estimate of the chemical and Nd isotope features of the metasedimentary country-rocks. The major and trace element data were obtained at the Centre de Recherche Pétrographique et Géochimique, Nancy, France, by inductively coupled plasma (ICP) atomic emission spectrometry and ICP mass spectrometry, respectively, following methods described by Carignan et al. (2001). The data are listed in Table 1. Sm-Nd isotope analyses were made in Clermont-Ferrand following isotope dilution, separation chemistry, and thermal ionization mass spectrometry methods described by Pin and Santos Zalduegui (1997). The results are given in Table 2, together with Nd isotope compositions corrected for in situ radiogenic decay of 147Sm, assuming geological ages of 570 and 600 Ma, respectively, and reported using the ϵ-notation. These results illustrate the range of initial isotope signature arising from a 30-Ma uncertainty in the true geological age. It can be seen that, in general, a ±30-Ma variation causes a shift in ϵ-value barely outside analytical uncertainty, highlighting that ϵNd values reported in Table 2 are not very sensitive to the uncertainty on the igneous emplacement age of the protoliths.
The previous geochemical studies of the Barrandian volcanic rocks (see Waldhausrová, 1997a,b, and references therein) have used a set of data for major and trace elements, some of which are potentially mobile during seawater alteration and low-grade metamorphism. Specifically, although the assessment of paleotectonic setting of the mafic metavolcanics was based on resistant elements, such as Ti, Mn, P (Mullen, 1983) and the lanthanides, the total alkali versus silica (TAS) classification scheme for volcanic rocks (Le Maitre et al., 1989) was used. However, Si, K, and Na, along with alkaline-earth elements (Ca, Sr, Ba), might have been redistributed to variable extent under the conditions to which the Barrandian rocks were subjected. Only elements extremely insoluble in aqueous fluids (Cox, 1995) and widely considered to be essentially immobile during zeolite- to green-schistfacies transformations (e.g., Wood et al., 1979; Saunders et al., 1980) have been considered in this work. These include high-field-strength elements (HFSE: Ti, Zr, Nb; e.g., Cann, 1970; Dungan et al., 1983); Th (Saunders and Tarney, 1984; Chen et al., 1986; Krauskopf, 1986); the rare earth elements (REE; Herrmann et al., 1974; Menzies et al., 1977, 1979; Dungan et al., 1983); and transition elements (Cr, Ni), including V (Shervais, 1982). However, significant enrichment of REE (particularly heavy REE [HREE]) and Th has been shown to have occurred in some extremely altered samples in which the HFSE remained essentially undisturbed (e.g., Wharton et al., 1995). For this reason, only least-altered material was selected for this study. Sm and Nd, adjacent elements of the chemically coherent lanthanide group, are involved in a radiogenic isotope system as parent and daughter isotopes, respectively, and so display a very high degree of chemical congruence. This congruence implies a very reduced relative mobility, thereby favoring a great resistance of the Sm-Nd isotopic pair with regard to geological disturbances (e.g., DePaolo, 1988). Indeed, it has been shown that seawater alteration is not able to fractionate Sm from Nd and disturb Nd isotope systematics (e.g., Staudigel and Hart, 1983), unless extremely high (on the order of 105) fluid/rock ratios are achieved (e.g., Jacobsen and Wasserburg, 1979). These conditions are presumably restricted to the immediate surroundings of hydrothermal vents and are not relevant to the samples investigated in this work. Based on these published lines of evidence, it is inferred that the chemical and Nd isotope features discussed in the following sections may be ascribed to the primary, igneous history of the studied rocks.
Immobile, incompatible trace element data are shown as chondritenormalized patterns, in order to depict their relative fractionation and highlight any “anomalous” behavior of certain diagnostic elements. Besides REE, these diagrams include the even more incompatible elements Th and Nb, which allows the degree of depletion of mantle sources to be assessed. In addition, the relative fractionation of Th, Nb, and light REE (LREE) is of great petrogenetic significance because it may provide evidence (as negative anomalies of Nb) for the involvement of a subduction component during the petrogenesis of ancient, altered basaltic rocks (e.g., Jenner et al., 1991). Indeed, supra-subduction zone lavas are almost invariably characterized by relatively high Th/Nb ratios (e.g., Sun, 1980) compared to mid-ocean ridge basalts (MORB; both of normal and enriched types) and within-plate basalts, which display nearly constant and very low values (∼0.05–0.08; e.g., Sun, 1980). Alternatively, elevated Th/Nb ratios in basaltic magmas may reflect merely contamination by wall-rock assimilation during magma ascent and emplacement (e.g., Dupuy and Dostal, 1984) and/or the addition of materials derived from the continental crust, because continental materials are characterized by a strong enrichment of Th and LREE relative to Nb (e.g., Taylor and McLennan, 1985). Nd isotopes provide a useful tool to assist the choice to be made between these alternative interpretations (e.g., Pin and Paquette, 1997), insofar as isotopically distinctive, old sialic crust was involved.
Several subgroups can be distinguished among the Barrandian metabasalts and further subdivided on the basis of their chondrite-normalized trace element patterns:
One subgroup gathers six samples (Klí-5, Teb-1, Lhv-2, Rou-1, Kruš-1, and Zb-1; Fig. 2A) with almost flat patterns (at ∼10× chondrite abundances) for moderately incompatible elements (specifically, elements from Nd to Lu, ranked as a function of increasing bulk partition coefficients during partial melting of upper mantle mineral assemblages). This subgroup has a strong relative depletion of the most incompatible elements (Th, Nb, La, Ce). In detail, the heavy REE patterns range from flat (Klí-5, Kruš-1) to slightly fractionated (Teb-1, Rou-1). All these samples display a distinct negative anomaly for Ti, combined in some cases (Kruš-1, Klí-5, Teb-1) with a faint negative anomaly of Zr. There is no significant Nb anomaly in these samples. Zr/Nb ratios range from 32 to 60 with an extremely high value (115) in sample Kli-5 (mean 58, standard deviation [SD] 30). These elevated ratios point to a strongly depleted mantle source (e.g., Le Roex et al., 1983). Overall, this group of samples, characterized by a strong depletion of the most incompatible trace elements, shows clear affinities with normal MORB (N-MORB), although the ubiquitous Ti anomalies, occasional Zr negative anomalies, and relatively low Ti/V ratios (range 20–30, mean 25, SD 4) are reminiscent of some suprasubduction zone influence. Such a subtle influence is corroborated by Th/Nb ratios (range 0.09–0.12, mean 0.10, SD 0.01) slightly higher than the value typical for N-MORB, ocean island basalts (OIB), and transitional magma types (∼0.07). Initial Nd isotope signatures (ϵNd600) calculated for these samples are all extremely radiogenic (from +8.0 to +9.3, mean +8.7, SD 0.4) and demonstrate that the mantle source was strongly depleted in Nd relative to Sm on a secular basis. In this respect, the mafic melts parental to the Barrandian “spilites” might have been extracted from a typical N-MORB mantle source, which was characterized, 600 Ma ago, by ϵNd values of approximately +9, according to the evolutionary model of Zindler (1982). In any case, the basaltic magmas did not suffer any significant interaction with ancient crustal materials characterized by unradiogenic Nd isotope composition (i.e., distinctly negative ϵNd values).
A second subgroup of six samples (Chř -1, Klí-1, La-2, Li-1, Reb-1, and UP-1) displays the same trace element characteristics, that is, flat chondrite-normalized patterns from Nd to Lu at ∼10 times the chondritic level, with negative Ti anomalies and strong relative depletion of the most incompatible elements from Th to Ce (Fig. 2B). However, relative abundances of Th tend to be higher, as shown by Th/Nb ratios ranging from 0.12 to 0.21 (mean 0.15, SD 0.03) and illustrated by faint negative Nb anomalies on chondritenormalized diagrams. Other wise, these samples share the same characteristics as the previous group, that is, elevated Zr/Nb ratios (34–82, mean 60, SD 20), moderate Ti/V ratios (19–27, mean 23, SD 3), and very radiogenic values (from +7.8 to +8.7, mean 8.3, SD 0.4). The sample with the deepest Nb anomaly (UP-1) also has a somewhat lower ϵNd value (+7.8), possibly suggesting that this sample suffered some crustal contamination by materials with high Th/Nb and low ϵNd, such as the gray-wackes forming most of the Barrandian Proterozoic (e.g., samples Nml-1 and Vis-1, with Th/Nb ∼0.60 and ϵNd600 approximately −1). However, sample Chr -1, with an identical ϵ Nd600 of + 7.8, does not display an elevated Th/Nb ratio (0.12), suggesting that mantle source heterogeneity might provide an alternative explanation for the observed range of ϵNd values.
Based on these data, it is apparent that there is no clear compositional gap between these two subgroups of samples. Indeed, the broad continuity of this set of samples is highlighted by multi-element patterns normalized to average N-MORB. In this diagram (Fig. 3), the samples display almost flat patterns from Nd to Lu, with a general relative depletion of Zr and Ti. For the most incompatible elements, their patterns continuously range from strongly LREE-depleted to very slightly LREE-enriched, but all show a marked negative Nb anomaly relative to La, and especially to Th. Such chemical features demonstrate that, although similar to N-MORB in many respects, these metabasalts are characterized by a relative depletion in HFSE, particularly Nb. For this reason, these samples are hereafter collectively referred to as “Nb-depleted basalts,” or simply, “depleted basalts.” Although this term is admittedly of limited significance, note that eight of the twelve samples in this group (i.e., Lhv-2, Zb-1, Klí-1, Klí-5, La-2, Reb-1, Rou-1, and Kruš-1) plot along a rather poor isochron (mean square of weighted deviations [MSWD] = 3.7) in a143Nd/144Nd versus 147Sm/144Nd diagram (not shown). This isochron indicates an age of 605 ± 39 Ma (uncertainty quoted at the 95% confidence level, using the Isoplot software; Ludwig, 1994), with an initial ϵNd of +8.8 ± 0.2, as calculated following the method of Fletcher and Rossman (1982). The four remaining samples (Chr -1, Li-1, Teb-1, and UP-1) plot along a linear array (MSWD = 0.5) with a similar slope and a lower initial ratio, but no meaningful age (598 ± 162 Ma) can be obtained because of the limited number of samples and the small spread of Sm/Nd ratios. Nevertheless, in the absence of better radiometric evidence, the imprecise date of 605 ± 39 Ma is considered to reflect the igneous emplacement of the depleted metabasalts, in broad agreement with the Ediacaran age suggested by micropaleonto-logical data for associated sediments.
A third subgroup comprising four samples (Lit-1, Lit-2, Zchl-1, and to some extent, UP-2) clearly depart from the previous rocks in showing chondrite-normalized REE patterns gently sloping from La to Lu, that is, a fractionation pattern opposite to that observed in the previous samples. Although negative Ti anomalies occur throughout (Fig. 4), the other HFSE abundances are in marked contrast with those of the previous samples. Indeed, Nb is much more abundant in these rocks and does not show any depletion, but instead an enrichment relative to adjacent elements on chondrite-normalized diagrams (Fig. 4), whereas Th contents are almost constant at ∼10 times chondrite. Zr/Nb ratios are much lower (12–18) than those measured in the previous, depleted, group and clearly suggest a different mantle source. Th/Nb ratios show very little scatter (0.09–0.10) and preclude significant degrees of crustal assimilation or recycling into the mantle source. Overall, the REE and HFSE features indicate derivation from a mildly enriched mantle. However, ϵNd600? values are still strongly positive (from +6.7 to +8.2), implying that the mantle source of these rocks was depleted in Nd relative to Sm on a time-integrated basis.
The last group of samples comprises five metabasalts (Boro-1, Kli-3, Kot-1, Mit-1, and Mit-2) and one felsic lava (“keratophyre,” Si-2) classified as a trachyte according to the TAS diagram (Le Maitre et al., 1989). All the metabasalts display strong enrichment in LREE, Th, and especially Nb, which gives rise to positive anomalies on chondrite-normalized diagrams (Fig. 5A). Along with very low Zr/Nb ratios (4–7) and, with two exceptions, high Ti/V (51–77), these geochemical features resemble those of alkaline basalts, extracted from strongly enriched mantle sources. Two samples (Boro-1 and Kot-1) show negative anomalies of Ti and Zr, associated with much lower Ti/V (25, 27) and interpreted to reflect fractionation of Fe-Ti oxide rather than being intrinsic to parental magmas. Indeed, the felsic derivative (Si-2, with 70 wt% SiO2) ascribed to the same group of samples has deep anomalies of Ti and Nb (and of Sr and Eu), suggesting that fractionation of a plagioclase and Fe-Ti oxide (plus pyroxene and/or amphibole) assemblage was involved (Fig. 5B). As in the previous group of samples, low values of Th/Nb (0.09–0.10) put strict limitations on the involvement, if any, of crustal assimilation or source contamination processes. ϵNd600 values measured in the mafic samples range from +3.8 to +5.1, whereas the metatrachyte has a slightly lower value (+ 3.2), which might allow for a subordinate amount of crustal assimilation during magma differentiation. Albeit significantly less radiogenic than those observed in the mildly enriched group of samples, the Nd isotope signatures of these alkaline basalts still reflect the time-integrated depletion of their mantle source.
The number of samples analyzed is far too limited to allow any sound characterization of metasedimentary rocks. Instead we simply note that the two samples of graywackes (Nml-1 and Viš-1, collected from two localities separated by a few hundred meters) have essentially identical incompatible trace element abundances (specifically, large Th and LREE enrichment, unfractionated HREE, large negative anomalies of Nb and Ti, and very small negative Eu anomaly; Fig. 6). Sm-Nd isotope characteristics, with ϵNd600 values close to zero, preclude major derivation from old continental crust sources but instead require a significant input of juvenile (volcaniclastic?) material, in agreement with the immature characteristics (e.g., Na2O > K2O) of these rocks (see also Jakeš et al., 1979; Drost et al., 2004). In comparison, the black shale sample Kruš-2 displays broadly similar chondrite-normalized patterns for Th, Nb, and LREE, but it has deeper anomalies of Eu and particularly Ti, and contains half the HREE. Its distinctly negative ϵNd600 value (−3.7) reflects a larger contribution to the finer-grained detrital supply of epiclastic component(s) with time-integrated LREE enrichment, probably derived from ancient sialic crust. Model ages relative to model-depleted mantle (DePaolo, 1988), interpreted to reflect the average crustal formation age of the mixture of detrital components present in the sample, are 1.1–1.2 Ga (graywackes) and 1.5 Ga (black shale), largely in excess of deposition age. This difference highlights the role of old, recycled components in the sedimentary input of the Neoproterozoic basin.
In summary, on the basis of chondrite- and N-MORB-normalized diagrams, selected ratios of immobile incompatible trace elements, and Nd isotope signatures, the Barrandian metabasalts may be ascribed to two major groups, extracted from contrasting mantle sources:
A depleted group, characterized by variably strong LREE-depletion, elevated Zr/Nb ratios (>30), distinct negative Ti anomalies, and highly radiogenic Nd isotopes (ϵNd600 from +7.8 to + 9.3). Variable Th abundances in this group of samples are reflected by the absence or presence of negative Nb anomalies on chondrite-normalized plots. However, multi-element patterns normalized to N-MORB all show well-defined negative anomalies of Nb, and to a lesser degree, Zr and Ti. This excess of LREE and Th relative to HFSE suggests magma generation from depleted mantle sources fluxed by hydrous fluids and probably also indicates silicate melts derived from an oceanic slab and subducted sediments. Relatively low Ti/V ratios are interpreted to reflect relatively oxidizing conditions compared to those prevailing at mid-ocean ridges (Shervais, 1982), possibly caused by higher water contents. The elevated ϵ Nd600 values preclude any significant contribution from old sialic crust, either as a contaminant during magma ascent and emplacement or as a component recycled into the mantle source by subduction.
An enriched group, comprising both a subgroup of samples mildly enriched in incompatible elements (Zr/Nb 12–18), and a subgroup of samples even more strongly enriched in LREE and especially Nb, with very low Zr/Nb (4–7) and elevated Ti/V ratios. These trace element features are typical of the within-plate style of mantle enrichment, related to the overall incompatibility of trace elements during mantle melting under relatively “dry” conditions. Although markedly radiogenic, Nd isotope signatures vary from elevated ϵNd600 (from +6.7 to +8.2) in the mildly enriched subgroup, to lower ϵNd600 values (from +5.1 to +3.8) in the strongly enriched subgroup, which includes a metatrachyte with ϵNd600 of +3.2. The decoupling observed between enriched trace element characteristics (e.g., LREE-enrichment, which requires an enriched source) and radiogenic Nd isotopes (which imply that the mantle source was depleted in Nd relative to Sm on a time-integrated basis) may suggest that the depleted source region was metasomatized shortly prior to the igneous event. Alternatively, recent magma mixing between depleted (MORB-like) and enriched melts could have been involved. In this case, mixtures containing between ∼60 and ∼80–90% of the depleted end-member would exhibit LREE enrichment while retaining positive ϵNd values (Anderson, 1982).
Interpretation of Geochemical Results
The results obtained in this study generally corroborate the conclusions of earlier geochemical investigations (e.g., Wald-hausrová, 1997a, and references therein), which revealed the presence of several distinct igneous suites among the Neoproterozoic low-grade metabasalts from the Barrandian area, specifically, an earlier tholeiitic series, that was followed by an alkaline series through a stratigraphically intermediate transitional series. Our new results, based on alteration-resistant trace elements and Nd isotope data, allow us to gain further insight into the mantle sources involved and to try to assess in a more detailed manner the geotectonic significance of these metabasalts.
First, Nd isotope signatures combined with Th/Nb systematics rule out any significant role for crustal assimilation during magma ascent through the crust. Indeed, highly radiogenic signatures are measured in most samples, irrespective of their degree of enrichment in Th, Nb, and LREE. There is no obvious negative correlation between ϵNd values and Th/Nb, as would occur in case of assimilation of old sialic crust with low time-integrated Sm/Nd ratios (i.e., low ϵNd) and high Th/Nb ratios. Even the LREE-enriched samples from the second major group have distinctly positive ϵNd values, and their Th/Nb ratios are very low (0.09–0.10), unlike what would be observed in case of significant contamination by continental crust or sediments derived therefrom. This distinction implies that the two major groups of basaltic magmas parental to the Barrandian metavolcanics were emplaced into extremely attenuated crust, or even into a purely ensimatic (i.e., oceanic) setting, and that most, if not all, of their variability in terms of highly incompatible trace elements and Nd isotopes reflects mantle source processes.
Second, trace element data emphasize a major, twofold distinction between contrasting types of mafic magmas that were extracted from different mantle sources, namely, a depleted group with supra-subduction zone affinities, and an enriched group resembling within-plate OIB.
The generally primitive, incompatible element–depleted group resembles N-MORB in several characteristics, especially LREE depletion and highly radiogenic Nd isotopes. Because they are all notoriously prone to remobilization during low-grade alteration and metamorphism, it is not possible to make a reliable use of largeion lithophile elements (LILE; alkalis and alkaline-earths), whose enrichment provides a sensitive monitor of the addition of a subducted component to a depleted mantle source (e.g., Saunders and Tarney, 1984). However, multi-element diagrams normalized to N-MORB display negative anomalies of Nb (Fig. 3). Bearing in mind that Nd isotope data preclude crustal contamination as a possible cause, this style of HFSE/REE and Th fractionation is believed to be diagnostic of magmas produced by partial melting of depleted mantle that was metasomatized and fluxed by a hydrous component derived from a subducting oceanic plate. Whether the metasomatizing agent, enriched in LILE but deficient in HFSE, was a hydrous fluid (e.g., Eiler et al., 1998), a silicic melt (e.g., Prouteau et al., 2001), or both, is an open question. However, low-degree partial melting of the upper part of eclogitized, subducted crust, including any sedimentary component, is likely to occur at depths of 125–175 km beneath back-arc basins (Sinton and Fryer, 1987). The very high ϵNd values measured in the samples with suprasubduction zone affinity (depleted group) put some constraints on the nature of subducted sediments, that is, their ultimate continental or oceanic provenance. In the present case, it is clear that any subducted sedimentary component should have been derived from a juvenile source, characterized by radiogenic Nd isotope signatures. This derivation would favor a model involving either a trench starved of sediments or containing mostly volcaniclastic sediments derived from an ensimatic arc.
Furthermore, the low concentrations, compared to those of N-MORB, of most “conservative” incompatible elements (i.e., those believed to be largely independent of the subduction component), such as Ti, Zr, Nb, HREE, and Y (Pearce and Peate, 1995), and their relative fractionation (e.g., high Zr/Nb or Yb/Nb ratios) suggest that the depleted group was generated from a N-MORB source mantle that was already depleted in incompatible elements, as commonly observed in intraoceanic arcs with active back-arc basins (Pearce and Parkinson, 1993).
The later-stage group of samples enriched in incompatible elements, particularly Nb and LREE, does not show any feature of subduction-related magmatism. Instead, this group has clear affinities with within-plate magmas, such as OIB, believed to be generated by a relatively low-degree of partial melting from enriched mantle sources. In very general terms, a popular model relates such magmas to very deep mantle sources connected to the surface by narrow ascending plumes (e.g., Hofmann, 1997), without direct relationship to plate tectonic processes. However, the geochemical data themselves can hardly provide unambiguous evidence on the ultimate depth of mantle reservoirs. Indeed, geological observations favor alternative models based on the occurrence of enriched domains fairly ubiquitous in the shallow mantle (including supra-subduction wedges; e.g., Morris and Hart, 1983; Gill, 1984) where they could form a so-called “perisphere” layer (Anderson, 1995). This enriched reservoir might consist of a relatively refractory mantle veined by enriched components, such as frozen low-degree partial melts extracted from the underlying mantle. The incompatible element composition of partial melts from such metasomatized sources would be dominated by vein materials (e.g., Wood et al., 1980). This enriched reservoir might be tapped wherever the overlying lithosphere fails under extensional stress, thereby allowing low-degree melts to egress (Anderson, 1995). Based on these concepts, it is inferred that the OIB-like magmas forming late-stage volcanic build-ups in the Neoproterozoic Barrandian basin correspond to relatively low-degree partial melts that are extracted, as a response to limited extension, from mantle sources that were not dominated by subduction-related components, but instead resembled enriched domains believed to be broadly ubiquitous beneath lithospheric plates.
Possible Geological Implications
The geochemical and Nd isotope data clearly favor an intra-oceanic supra-subduction zone setting for the earlier, depleted group of the Barrandian metabasalts. More specifically, the relatively minor contribution of subduction-related component in several samples otherwise similar to N-MORB is reminiscent of extensional back-arc basins, which are commonly floored by MORB and basalts (commonly referred to as “back-arc basins basalts” (BABB; Fryer et al., 1981) that are geochemically transitional between MORB and island arc tholeiites (e.g., Saunders and Tarney, 1984; Volpe et al., 1987). Indeed, a back-arc or interarc rift environment would be in keeping with the thick pile of submarine clastic sedimentary rocks (mostly turbiditic graywackes) associated with these early-stage volcanics, and this geodynamic setting has already been advocated for a long time (see references in Křibek et al., 2000). In terms of modern analogue, the depleted metabasalts might correspond to magmas emplaced into intra-oceanic back-arc basins, such as the Lau basin (e.g., Hawkins, 1995; Pearce et al., 1995) or the Sumisu rift in the Izu-Bonin arc (e.g., Hochstaedter et al., 1990), where very high sediment accumulation rates (up to 4 km/Ma) have been reported (Marsaglia et al., 1995). The incipient, rifting stage of back-arc basin opening is commonly accompanied by the emplacement of a bimodal association of basalts and Na-rich felsic lavas, as exemplified by the nascent Sumisu rift (Hochstaedter et al., 1990) or the Northern Mariana Trough (Gribble et al., 1998). In contrast, only basaltic magmas are erupted during the more mature spreading stage, which may begin when the back-arc basin is 100–150 km wide (Gribble et al., 1998). The lack of significant amounts of felsic lavas in the studied area might therefore suggest that the basaltic magmas of the depleted group were formed by decompression melting in a relatively mature extensional setting. Moreover, by analogy with the systematic variation of the composition of the subduction-related component with the distance to the arc observed in the Lau basin (Pearce et al., 1995), the clear Th signal documented in the Barrandian samples (Fig. 3) might be tentatively interpreted to reflect generation relatively close (∼50 km or less) to the arc.
Although an ensimatic back-arc basin setting is favored by geochemical features of volcanic rocks, the preliminary Nd isotope data obtained on sedimentary rocks (e.g., model ages far in excess of deposition ages) indicate a substantial contribution from continental crustal components, mixed with a juvenile, presumably arc-derived component. Indeed, a few 2.0-Ga-old grains were found among detrital zircons of a sample of gray-wacke from the upper part of the Neoproterozoic Barrandian, with a maximum deposition age of 564 ± 16 Ma (Drost et al., 2004). Although these 2-Ga grains might have suffered multiple sedimentary recycling and do not necessarily imply a direct provenance from Paleoproterozoic basement, their presence and the scarce Nd isotope data available suggest that the intraoceanic arc/back-arc system inferred from the chemical and isotopic signature of igneous rocks was not far removed from a continental land mass, at least during the latest Neoproterozoic.
Based on major and trace element discrimination diagrams, Drost et al. (2004) favored a back-arc setting of deposition and a continental island arc provenance for Neoproterozoic gray-wackes. A composite provenance, involving juvenile volcaniclastic components from an oceanic island arc on the one hand, and epiclastic components from continental source(s) on the other hand, might provide an alternative interpretation.
The association (e.g., at Mitov and Koterov) of enriched, OIB-like pillow-lavas with sediments deposited in an inter- or supratidal environment (Pouba et al., 2000) shows that the late-stage, enriched magmas were able to build significant volcanic edifices, forming seamounts and even islands during the late evolutionary stage of the “spilitic series” of the Barrandian Neoproterozoic. Basalts containing a nonsubduction-related component, enriched in Nb, Zr, and LREE and geochemically similar to E-MORB or OIB, are not rare in recent arc and back-arc settings (e.g., Morris and Hart, 1983; Gill and Whelan, 1989). Examples from back-arcs include the North Fiji Basin (Price et al., 1990), the Sumisu rift (Hochstaedter et al., 1990), and particularly the Japan Sea, where mildly alkaline and alkaline basalts form seamounts and volcanic islands (e.g., Pouclet et al., 1995). The source heterogeneity documented by these rocks has been variably interpreted to reflect either veins or blobs of enriched mantle contained in a more depleted matrix (e.g., Wood et al., 1980), or injection of new mantle. The first class of models may account for the generation of enriched magmas during the early stages of rifting, through preferential melting of shallower enriched domains as a result of higher geothermal gradients. In contrast, the occurrence of enriched magmas at a late-stage, and the switch from early subduction-related to younger within-plate magmatism, would rather favor the alternative hypothesis invoking injection of new mantle. Vanishing or even missing subduction-related components are indicated in the late-stage, enriched Barrandian metabasalts. This dearth in turn suggests that the underlying upper mantle was no longer fluxed by fluids and/or melts from a descending slab and that mantle flow allowed injection of new, more enriched material in the melt source, replacing the former hydrous mantle wedge. Similar changes in recent arc and back-arc systems have been interpreted to reflect “plume mantle” flowing from beneath the subduction hinge into the back-arc region, either laterally around the edges or through gaps (tears) of retreating subducting slabs (Schellart, 2004, and references therein). However, the mantle flow compensating for the retreat (roll-back) of subducting slab may also involve enriched mantle dragged from the base of the overriding plate toward the back-arc basin (e.g., Martinez and Taylor, 2002). In both cases, mantle flow is associated with the retrograde motion of a subduction hinge (roll-back) that occurs as a natural consequence of the negative buoyancy of sufficiently old subducting oceanic slabs with regard to the surrounding mantle (Elsasser, 1971). For the same reason, an extensional regime commonly prevails in the overriding lithospheric plate (Hamilton, 1995). In the Barrandian case, such a retreating slab scenario could account for (1) the inferred, strong extension of the overriding plate, with opening of a basin, accompanied and followed by BABB magmatism and the deposition of thick clastic sediments; (2) the extinction of the supra-subduction zone geochemical component; and (3) the switch to enriched, within-plate style magmatism.
An alternative interpretation would postulate not a mere ocean-ward migration of the subduction zone, but its death. For example, this extinction might have occurred as a result of impingement of a spreading ridge with the intraoceanic trench, leading to mutual annihilation and switch to a transform plate boundary. This process would be accompanied by the opening of a “slab window” beneath the upper plate of the extinct subduction zone, which would allow for the upflow of new mantle material in the former arc and back-arc system, thereby accounting for the generation of late-stage, enriched basalts. In the absence of precise chronological and tectonic constraints, it is not yet possible to favor any of these highly tentative interpretations.
Comparison with Other Late Proterozoic Segments
The closer example of relatively well-preserved Precambrian formations occurs in the Brno Massif, which corresponds to the exposed part of the largely covered Bruno-Vistulicum basement block (Fig. 1, inset). On the basis of geochemical and Sr-Nd isotopes, three contrasting crustal blocks (“terranes”) were recognized (Finger and Pin, 1997; Finger et al., 2000a). The eastern Slavkov terrane consists mainly of ca. 590-Ma quartzdiorites, tonalities, and granodiorites with relatively primitive isotope signatures (87Sr/86Sri ∼0.704–0.705; −3 < ϵNdi < −1). In marked contrast, the western Thaya terrane displays K-rich granitoids, also emplaced ca. 580 Ma, but characterized by crustal isotopes (87Sr/86Sri ∼0.708–0.710; −7 < ϵNdi< −4). The intervening Brno-Breclav terrane consists of a fault-bounded belt of volcanic and plutonic mafic rocks that probably corresponds to a suture. A 725 ± 15-Ma Pb/Pb zircon evaporation age was measured on a subordinate metarhyolite (Finger et al., 2000b), interpreted to be cogenetic with the overwhelming mafic rock-types based on similar ϵNd725 values of +6.8.
Recent intraoceanic subduction systems may reach very great cumulative length (e.g., some 2500 km along the Izu-Bonin-Mariana or the Tonga-Kermadec arc systems). This observation suggests that even larger-scale correlations throughout Cadomian Europe should be attempted.
As long recognized, the Lower Brioverian series from northern Armorican Massif (northwest France) show similarities with the Central Barrandian domain. Sedimentary rocks consist of monotonous terrigeneous sediments with mixed volcaniclastic and reworked continental provenance (Dabard, 1990). A typical feature is the frequent occurrence of interbedded black cherts that are interpreted to represent silicified terrigeneous and evaporitic deposits (Dabard, 2000), as also inferred in the Barrandian. Concerning igneous rocks, an episode of intra-arc or back-arc extension is documented ca. 610 Ma, as exemplified by (1) the Paimpol spilites (containing minor rhyolites dated 610 ± 9 Ma) with arc tholeiite affinities (Egal et al., 1996) and ϵNd of approximately + 6 (Dabard et al., 1996); (2) the Erquy spilitic series, dated 608 ± 7 Ma (Cocherie et al., 2001) and the broadly equivalent Lanvollon bimodal suite, with still rather uncertain geodynamic settings (back-arc basin according to Cabanis et al., 1987; within-plate rift according to Lees et al., 1987, and Egal et al., 1996); and (3) the Yffiniac-Belle-Isleen-Terre gabbros, emplaced 602 ± 8 Ma and 602 ± 4 Ma, respectively (Guerrot and Peucat, 1990). Certain amphibolites associated with these gabbros were reported to be chemically similar to oceanic tholeiites (Chantraine et al., 2001). Other scattered mafic volcanics occur within the Lower Brioverian series of the Lamballe and St. Lô formations, composed of terrigeneous sediments containing interbedded black cherts. The mafic volcanics show either a within-plate alkaline affinity in the Lamballe Formation (Cabanis et al., 1987) or a similarity with N-MORB in the St. Lô Formation (Dupret et al., 1990). Close to the southeastern end of Armorican Massif, the Precambrian Mauges Group is composed of a several-kilometerthick monotonous sequence of clastic rocks ranging from quartz-feldspar metagray-wackes to metapelites, together with black cherts and graphitic schists. Mafic metaigneous rocks (gabbros, lava flows, breccias, dikes, and cinerites) display geochemical features similar to those of transitional tholeiitic basalts emplaced in a within-plate extensional setting (Cabanis and Wyns, 1986).
In summary, a broadly extensional tectonic regime is clearly indicated by ca. 610–600-Ma mafic or bimodal volcanic suites, with both back-arc and within-plate geochemical affinities. However, Nd isotope data are too scarce to monitor the degree of contamination by crustal materials or to assess the ensialic versus ensimatic setting of the Lower Brioverian basin(s). This episode was bracketed by arc-related plutons dated from ca. 750 Ma to ca. 625 Ma (with strongly radiogenic Nd isotope signatures; Samson et al., 2003), and ca. 580-Ma syn- to late-kinematic diorite intrusions (e.g., Nagy et al., 2002, and references therein) with negative ϵNd values, interpreted to reflect generation in an active continental margin containing an ancient basement, such as the 2.1-Ga relics (Icartian) documented in that region.
The Sierra Morena (southwest Spain) exhibits Precambrian rocks variously overprinted by Cadomian and Variscan metamorphisms and deformations (e.g., Quesada, 1990; Eguiluz et al., 2000). Among these formations, the Serie Negra consists of a >5-km-thick succession of graphite-rich, turbiditic metapelites and metagraywackes containing laminated black cherts and marbles (Quesada, 1990). This series has been subdivided into a lower group (Montemolin Group) made of thin-bedded quartz-rich graywackes and graphite-rich pelites with black cherts and local marbles, containing abundant amphibolites in its upper part, and an upper one (Tentudia Group), consisting of progressively thicker beds of massive graywackes that contain an increasing proportion of calc-alkaline volcanic clasts. The maximum age of deposition of the Tentudia group is constrained by the 564 ± 30-Ma age of the youngest detrital zircons dated in a metagray-wacke (Schäfer et al., 1993). The youngest detrital grains measured in a graphite-rich metapelite from the Montemolin Group gave an average SHRIMP age of 591 ± 11 Ma (Ordoñez Casado, 1998). Ages of 600 and 610 Ma have been reported for the igneous protoliths of amphibolites from the Montemolin Group and the Badajoz-Cordoba shear zone, respectively (Schäfer, 1990, as cited in Bandres et al., 2002, 2004). Sm-Nd analysis of a sample from these broadly tholeiitic rocks gave an ϵNd600 value of +7.4 (Ordoñez Casado, 1998), indicative of a mantle source with strong time-integrated depletion of LREE. In the absence of more detailed geochemical studies, it is not possible to infer whether these rocks reflect subduction- or nonsubduction-related extensional setting. However, the sedimentary successions bear similarities with Lower Brioverian formations of Armorican Massif and with the Central Barrandian graywackes. Moreover, dioritic plutons of probable continental arc affinity were emplaced ca. 580 Ma (Merida Massif; Bandres et al., 2004), a situation reminiscent of that documented in northern Armorican Massif and the Brno Massif. In the southeast of the Ossa-Morena zone, conglomerates, arkoses, and shales of lowermost Early Cambrian age unconformably overlie the San Jeronimo Formation, composed of andesites interbedded with conglomerates, sandstones, and lutites, with a minimum thickness of 1 km. Based on micro-fossil evidence and stratigraphic context, the San Jeronimo Formation was ascribed to the Varangerian glacial stage (Quesada et al., 1990), possibly indicating a ca. 580-Ma age by correlation with the Gaskiers Formation glacial deposits in Newfoundland (Bowring et al., 2003). The andesites show typical calc-alkaline chemistry and strongly radiogenic Nd isotope signatures (Pin et al., 2002), implying a time-integrated depleted mantle source. However, sedimentary rocks interbedded with the andesites have negative ϵNd values indicative of continental crust sources, which preclude a purely intraoceanic arc setting. These combined pieces of evidence suggest generation in a supra-subduction environment located on relatively juvenile crust, such as an island arc previously accreted to a continental margin.
This short review shows that subduction-related arcs and back-arc basins were a general feature throughout the European domain during the 750–580-Ma period, as part of a much wider domain extending from present-day northwest Africa (Morocco) and Avalonia to the Arabo-Nubian shield (Fig. 7). The opening and spreading of back-arc basins ca. 610 Ma is suggested to have occurred in all three European examples, but whether only ensimatic or both ensialic and ensimatic basins were involved is still an open question, given the lack of adequate geochemical data. These basins acted as efficient traps for detrital sediments derived from both juvenile arcrelated and recycled continental sources that played a significant role in local crustal growth. Overall, this picture is reminiscent of modern arc and back-arc systems from the western Pacific region, where large intraoceanic subduction systems fringe the major continental masses of the Asian and Australian plates, along with a complex mosaic of microplates and magmatic arcs that include intervening basins floored either by oceanic crust or attenuated continental crust. It is speculated that the Neoproterozoic paleogeography of the Cadomian realm broadly resembled such patterns.
Based on combined Nd isotope and trace element evidence, the spilitized basalts of the Central Barrandian Neoproterozoic reflect contrasting magmas extracted from fairly different mantle sources. During an earlier stage (tentatively dated at 605 ± 39 Ma based on a whole-rock Sm-Nd isochron), basalts broadly similar to N-MORB but showing negative anomalies of HFSE were extracted from a source that was strongly depleted in LREE on a time-integrated basis and emplaced as lava flows in a strongly subsiding sedimentary basin. Subsequently, Nb-rich basalts extracted from enriched mantle sources were emplaced as shallow to subaerial volcanic edifices. This kind of evolution of mafic magmatism is reminiscent of some recent intraoceanic back-arc basins, where a switch from supra-subduction zone to within-plate–like magmatism is documented. This change might have occurred either simply because of ocean-ward migration of the subduction zone or as a result of impingement of a spreading ridge with the intraoceanic trench, leading to mutual annihilation and evolution to a transform plate boundary. Even though the geochemical affinities of igneous rocks clearly favor an ensimatic setting, the presence of a thick sedimentary pile containing both juvenile and recycled old crustal components suggests that the inferred Late Proterozoic intraoceanic arc and back-arc system was located near a continental mass.
APPENDIX: SAMPLE LOCATIONS
Kli-5: Metabasalts (tholeiitic), outcrops of a massive flow along the road N of the Klíc ava dam, north of Zbec no. Main central volcanic belt in the Rakovník area.
Zb-1: Metabasalts (tholeiitic), from the Zbečno quarry, opposite the railway station. Main central belt in the Rakovník area.
Rou-1: Porphyritic basaltic meta-andesite, outcrops on the Rou-pov castle hill (southwest of Přeštice). Main central belt in the Přeštice sector.
Kruš1: Metabasalts (tholeiitic), massive flow in the Krušec quarry near Chudenice village. Main central belt, in the Klatovy area.
Teb-1: Metabasalt, massive flow in a small abandoned quarry east of Třebobuz village, southwest of Všeruby. Stříbro volcanic belt.
Lhv-2: Metabasalt, outcrops of a massive flow on top of the hill northeast of Luhov village, southwest of Všeruby. Str íbro volcanic belt.
Kli-1: Metabasalt from an abandoned quarry, at the northern end of the Klíčava dam, north of Zbeřno. Main central belt in the Rakovník area.
La-2: Basaltic meta-andesite, outcrops of a massive flow on the hill Čihadlo, south of Lány village. Main central belt in the Rakovník area.
UP-1: Basaltic pillow lava, “Kněžská skála,” rocky outcrops on the left bank of the Berounka River near Nezabudice village, north of Skryje. Main central belt in the Hřebečníky sector.
UP-2: Basaltic pillow lava, “Čertova skála,” rocky outcrops on the left bank of the Berounka River near the Týr ovice village, north Skryje. Main central belt in the Hr ebec níky sector.
Chř-1: Metabasalt (transitional), outcrops of a massive flow in the vicinity of Chříč village, northwest of Zvíkovec. Main central belt in the Hřebečníky sector.
Li-1: Vitrocrystalloclastic, basaltic laminated tuff, outcrops from the locality “Liška” in the vicinity of Poleň village. Main central belt in the Klatovy area.
Reb-1: Metabasalt, outcrops of a massive flow near Řebří village, in the vicinity of Svojšín village, west of Stříbro. Svojšín volcanic belt.
Lit-1: Metabasalt, massive flow from the Litice quarry, south of Plzen. Main central belt in the Plzeř area.
Lit-2: Actinolitized basalt in the Litice quarry.
Zchl-1: Metabasalt, outcrops of a massive flow near Záchlumí, NNW of Stříbro village. Svojšín volcanic belt.
Kli-3: Mugearite, outcrops at the northwest end of the Klíčava dam north of Zbečno village. The main central belt in the Rakovník area.
Kot-1: Metabasalt from the abandoned quarry near Koterov village, southeast of Plzeň town. Main central belt in the Plzeň area.
Boro-1: Transitional basalt, small outcrops near Borovno village, northeast of Blovice town. The southern volcanic zone in the Blovice sector.
Mit-1: Alkali basalt, massive flow from the quarry west of Mítov village, east of Blovice town. Southern volcanic zone in the Blovice sector.
Mit-2: Alkali basalt, pillow lava, from the Mítov quarry.
Si-2: Devitrified glassy trachyte, massive flow from the Slatina quarry, southwest of Rakovník town. Slatina-Pavlíkov strip.
Viš-1: Graywacke, outcrops on the left bank of the Berounka River, at the locality Višňová near Roztoky u Křivoklátu. Main central belt in the Hřebeč níky sector.
Nml-1: Graywacke, outcrops on the left bank of the Berounka River, Nezabudice mill near Nezabudice village. Main central belt in the Hřebečníky sector.
Krus-2: Black shale from an intercalation in basaltic flows, Krušec quarry near Chudenice village. Main central belt in the Klatovy area.
This study was supported by travel grants to one of us in the scope of the Czech-French cooperation (Barrande Project). We are grateful to Mr. Dašek for drawing some of the figures. Constructive reviews of the manuscript by Dr. R. D'Lemos and Dr. S. Samson are gratefully acknowledged. This article is a contribution to the International Geological Correlation Program Projects 453 and 497.
Figures & Tables
The Evolution of the Rheic Ocean: From Avalonian-Cadomian Active Margin to Alleghenian-Variscan Collision
- alkali basalts
- Barrandian Basin
- Bohemian Massif
- Central Europe
- chemical composition
- chemical ratios
- Czech Republic
- igneous rocks
- island arcs
- isotope ratios
- metaigneous rocks
- metamorphic rocks
- plate tectonics
- rare earths
- stable isotopes
- trace elements
- upper Precambrian
- volcanic rocks
- Uslava Group
- Zvikovec Group
- Rakovnik Group
- Rabstejn Group